This Schizophrenia Outcome Survey compared medical costs, psychopathology and adverse events in outpatients for 2years following hospitalisation for an acute schizophrenic episode.

Adults stabilised with haloperidol, olanzapine or risperidone entered this observational study ≤1month after discharge and were assessed at baseline, 3, 6, 12, 18 and 24months using Brief Psychiatric Rating Scale (BPRS), Clinical Global Impression (CGI), Global Assessment of Functioning and adverse events reporting.

Among 323 patients (haloperidol 32, olanzapine 149, risperidone 142), baseline characteristics were similar in the olanzapine and risperidone groups, except for more first episodes in the risperidone group (p=0.01). Haloperidol patients were more often single and institutionalised, less educated, had more residual schizophrenia, were longer hospitalised in the previous year, took more corrective and psychotropic drugs and had more extrapyramidal symptoms (EPS) and gynaecomastia (all significantly). Sixty-eight percent of patients completed a 2-year follow-up. In all groups, CGI and GAF improved during the first 3months (both p<0.0001) while BPRS deteriorated in the first year (all within group changes p<0.05, between group changes NS) before it stabilised. There were no significant differences in hospitalisations and no change in social profile. At the last visit, 66% of haloperidol (p<0.01), 35% of olanzapine (NS) and 39% (NS) of risperidone patients had ≥1 EPS; 69% (p<0.013), 40 and 44%, respectively, had ≥1 sexual problem (NS). Mean weight gain was 0.4 (NS), 2.6 (p<0.05) and 2.6kg (p<0.05), respectively.

In this naturalistic study, treatment allocation might have introduced a bias in the interpretation of efficiency results, but olanzapine and risperidone caused less EPS than haloperidol during 2years of outpatient follow-up.